Inference for the cost-effectiveness acceptability curve and cost-effectiveness ratio

Area of Science:

• Health economics
• Biostatistics
• Clinical trial analysis

Background:

• Cost-effectiveness (C/E) analysis is crucial for healthcare decision-making.
• Existing Bayesian methods for C/E acceptability curves often oversimplify variance estimation.
• Frequentist inference for C/E ratios can yield unreliable results.

Purpose of the Study:

• To present simplified Bayesian and frequentist inference methods for C/E acceptability curves.
• To address limitations in current Bayesian approaches, particularly regarding variance uncertainty.
• To offer a more robust alternative to standard frequentist C/E ratio inference.

Main Methods:

• Development of a simple Bayesian computation for the C/E acceptability curve.
• Introduction of a frequentist analogue to the Bayesian method.
• Utilizing distributional assumptions that account for uncertainty in estimated variances.

Main Results:

• The proposed Bayesian computation is simpler and more accurate than existing methods.
• The frequentist analogue provides a reliable alternative for C/E ratio inference.
• The methods correctly incorporate uncertainty in variance estimation, unlike normal posterior assumptions.

Conclusions:

• The presented methods offer improved accuracy and simplicity for Bayesian and frequentist C/E acceptability curve analysis.
• Accounting for variance uncertainty is critical for reliable cost-effectiveness inference.
• A Bayesian interval is recommended over standard frequentist procedures for C/E ratio inference.